Workshop on
“Monetary policy: Recent experience and future directions”
Could monetary policy
have helped prevent the financial crisis?
Lorenzo Bini Smaghi
Member of the Executive Board, European Central Bank
Bank of Canada, Toronto 9 April 2010
Introduction1
There is a broad consensus that the financial crisis has been caused by several factors, such
as: failures in financial regulation and supervision; structural changes in the financial sector,
including the increased importance of the shadow banking system and securitisation; and
global macro-economic imbalances. These factors, together with widespread financial market
integration, which in turn creates greater scope for contagion, have been particularly relevant
even in countries where credit growth was relatively contained and real estate did not
experience a boom.
There is much less consensus on whether monetary policy could have helped prevent the
financial crisis, or, more provocatively, on whether errant monetary policy caused the
financial crisis. These are uncomfortable questions for central bankers – but ones we cannot
avoid. Only by confronting such challenges honestly can we hope to understand the crisis,
and thereby guard against its repetition.
Some take the view that monetary policy played, at most, only a marginal role. In this
account, failures of financial regulation and supervision – rather than of monetary policy –
lay at the heart of the crisis.2 Yet an influential body of opinion takes an opposing line,
claiming that an overly accommodative monetary policy – at the global level, and perhaps
especially in the United States – was among the key causes of the crisis.3 Its advocates argue
that short-term interest rates were kept “too low for too long” following the bursting of the
dot-com bubble at the turn of the century, and led to the economic imbalances that ultimately
threatened the financial and macroeconomic stability of the world economy.
The truth probably lies somewhere between these two extremes. No doubt financial
regulation and supervision were weak. No doubt price stability is necessary but insufficient
per se to achieve financial stability. Yet there is also no doubt that a monetary policy which
turns out to be too lax to achieve price stability is likely to be responsible for fuelling
excessive credit growth and thereby creating the potential for financial instability. There is
substantial literature explaining this relationship, on which I will not elaborate further.4
1
I wish to thank José Luis Peydró-Alcalde and Huw Pill for their contribution to preparing the speech. I
remain sole responsible for the opinions.
2
See: B. Bernanke (2010), “Monetary policy and the housing bubble,” speech at the 2010 AEA meetings
http://www.federalreserve.gov/newsevents/speech/bernanke20100103a.htm.
3
For a response to Bernanke (2010), see: J. Taylor (2010), “The Fed and the crisis: A reply to Ben Bernanke,”
Wall Street Journal (11 January).
4
See, inter alia: C. Borio and H. Zhu (2008), “Capital regulation, risk-taking and monetary policy: A missing
link in the transmission mechanism?” BIS working paper no. 268; G. Jiménez, S.R.G. Ongena, J-L. Peydro
1
Upside threats to price stability were certainly emerging prior to the crisis, on the back of
surging commodity prices and strong global growth. Let’s not forget, for example, that US
headline inflation increased from 2.3% on average in 2003 to 3.2% in 2006 and 3.8% in 2008
(See figure 1). Core inflation increased from 1.5% in 2003 to 2.5% in 2006 and 2.3% in 2008
(See fig. 2). Euro area annual inflation reached 3.3% on average in 2008.5 To contain these
upside risks to price stability, a case can be made that interest rates should have been higher
than was the case before the crisis.
Central to a honest self-evaluation of the role played by monetary policy is our adoption of a
“real-time” perspective. The benefits of hindsight are enormous – but also potentially
misleading. If we had known ahead of time that Lehman Brothers would fail in September
2008, we would no doubt have taken different policy decisions in the preceding months. But
such considerations are not a meaningful guide to future policy. By their nature, economic
“shocks” cannot be foreseen.
Taking these aspects into account, it would be a mistake to think that the way monetary
policy was conducted did not play any role in the outbreak of the financial crisis. Monetary
policy can actually be considered one of the factors that, in some countries, contributed to the
crisis, at least to the extent that it was too accommodative to maintain price stability. The
overly loose global monetary policy may have weakened the anchoring of price stability and
at the same time unleashed a wave of euphoria and excessive risk-taking in the financial
sector.
It would be a waste if this financial crisis did not give us the opportunity to think deeply
about the framework for monetary policy-making – the objective of policy; the models
underpinning our analysis; and the indicators on which we focus when taking policy
decisions. Even if excessively loose monetary policy is only partly to blame for the financial
crisis, a case can be made that it fuelled the accumulation of financial imbalances that
underlay the crisis. This prompts us to ask: what were the flaws in the decision-making that
resulted in – according to some observers – policy errors? And, more importantly, how can
we correct these flaws – and avoid any repetition – in the future?
and J. Saurina Salas (2010), “Credit supply: Identifying balance sheet channels with loan applications and
granted loans,” CEPR discussion paper no. 7655; Y. Altumbas, L. Gambacorta and D. Marques (2010),
“Does monetary policy affect bank risk-taking?” ECB working paper no. 1166; and A. Maddaloni and J-L.
Peydró (2009), “Bank risk-taking, securitisation, supervision and low interest rates: Evidence from lending
standards,” ECB working paper, forthcoming.
5
On the basis of the harmonised index of consumer prices (HICP).
2
Since the seminal contribution of Theil,6 policy-making has been characterised as a “control
problem”. Such a characterisation embodies three elements: clearly defining the policy
objectives; articulating a model of the economy by which policy instrument settings and
economic shocks are mapped into outcomes; and defining the information structure facing
policy-makers. This last element determines the extent to which they can identify shocks and
structural changes in real time.
Using this approach as an organising framework, we can re-phrase our question: if central
banks made mistakes in the run-up to the financial crisis, was this because they had the
wrong objectives? Or used the wrong model? Or misread the conjunctural situation? Or were
the errors a result of the combination of these three factors?
1. The objective of monetary policy
A large body of academic literature, stemming from the insights of Friedman and Phelps,
provides the theoretical basis for establishing price stability as the objective of monetary
policy. Such an approach found overwhelming empirical support in our experience of the
Great Inflation during the 1970s. It has reached its apogee in the modern benchmark
macroeconomic model, which articulates – on a state-of-the-art, micro-founded basis – the
welfare costs of deviations from price stability.7
Such thinking had an important bearing on the design of the institutional framework for
Monetary Union in Europe. The ECB has been assigned the primary objective of maintaining
price stability in the euro area, establishing a clear hierarchy of goals with price stability at
the fore. The clarity of this objective bolsters the credibility of the single monetary policy: it
supports the institutional independence of the ECB and helps to stabilise private longer-term
inflation expectations at levels consistent with price stability.
At some other central banks, monetary policy has been given a dual mandate, with objectives
such as economic activity, employment, financing conditions or financial stability in addition
to the goal of price stability. I do not want to give the impression that we are indifferent to
such concerns at the ECB: on the contrary, I will discuss in a moment how we can help to
6
See H. Theil (1964). Optimal decisions for government and industry, Amsterdam: North Holland.
7
See: M. Goodfriend and R. King (1997), “The new neoclassical synthesis and the role of monetary policy,”
NBER Macroeconomics Annual 12, pp. 231-296; and J.J. Rotemberg and M. Woodford (1999), “Interest
rate rules in an estimated sticky price model,” in ed. J.B. Taylor Monetary policy rules, Chicago University
Press, pp. 57-126.
3
attain them. But monetary policy should not be over-burdened with additional objectives. It is
too blunt an instrument to be effective in achieving all the goals I have just listed.
For example, a central bank with a dual mandate to maintain price stability and support
employment faces a particular dilemma in the context of a so-called “jobless recovery.” Its
ability to act in a timely manner to contain emerging inflationary risks may be curtailed by its
obligation to support employment. Moreover, since labour market developments are lagging
indicators of the cycle, a central bank assigned an explicit employment objective may
systematically tighten monetary policy later and more cautiously than necessary to maintain
price stability. Such behaviour could impart an “inflation bias” to the economy.
These factors run counter to recent proposals to assign monetary policy additional financial
stability objectives.8 Rather, maintaining price stability over the medium term – an important
addendum, to which I will return – is the appropriate objective of monetary policy. Given its
neutrality over the longer run, monetary policy’s ability to pursue other objectives at that
horizon is heavily circumscribed. And by maintaining price stability, monetary policy creates
an environment conducive to financial stability, economic growth and employment creation.
If price stability is the appropriate objective of monetary policy, how should it be quantified?
At the ECB, we have defined price stability as annual consumer price inflation of “below, but
close to, 2%”. Whether explicit or implicit, similar targets have been established by central
banks throughout the world.
Recently, proposals have been made to raise inflation objectives substantially – more than
doubling them in the euro area case.9 Frankly, these proposals are foolhardy. Any shift in
inflation objectives at times of economic stress invites the charge of opportunism: if you are
willing to raise inflation objectives from 2% to 4%, then why not 6% or 10%? Such thinking
puts at risk the hard-won credibility of the existing monetary policy frameworks. To the
extent such shifts are motivated on public finance grounds, they suggest that monetary policy
is subordinate to fiscal concerns. And concerns about deflation risks should be addressed by
refining the conduct of monetary policy, rather than redefining its objective.
In particular, it is crucial that the objective of price stability is pursued symmetrically. Of
course, asymmetries may exist in the structure of the economy. Examples include: any lower
8
See: e.g. S.S. Roach (2010). “The post-crisis fix: Regulatory or monetary policy remedies?” paper presented
at the Reserve Bank of India’s First International Research Conference (February); P. De Grauwe and D.
Gros (2009), “A new two-pillar strategy for the ECB,” CEPS policy brief no.191.
9
See: O. Blanchard, G. Dell’Ariccia and P. Mauro (2010). “Rethinking macroeconomic policy,” IMF staff
position note no. 10/03 http://www.imf.org/external/pubs/ft/spn/2010/spn1003.pdf.
4
bound on nominal interest rates; and an aversion to falls in nominal wages. These need to be
taken into account in policy making. But this does not imply that monetary policy should
demonstrate greater aversion to deflation rather than inflation (or vice versa). An excessive
and asymmetric fear of deflation explains in part why interest rates may have been kept “too
low for too long” prior to the financial crisis. Pursuing price stability in a symmetric manner
would guard against the danger of repeating this mistake.
To sum up, I believe that the recent financial crisis has not challenged the principle according
to which the goal of monetary policy should be price stability. In practice, this is recognised
by most leading central banks. At the ECB, we are fortunate that the primacy of price
stability has been made explicit.
With this in mind, it is difficult to attribute any pre-crisis “policy error” to the pursuit of the
wrong monetary policy objective. Nonetheless, we are already witnessing the emergence of
siren calls to change the objective – in particular, for monetary policy to target a higher
steady-state inflation rate, difficult to reconcile with price stability. These calls must be
resisted.
Emphasising the primacy of price stability objective over the medium term does not imply
central bankers should be, in the words of Mervyn King, “inflation nutters”.10 Using Theil’s
framework to characterise the monetary policy problem also helps to clarify its dynamic
nature. Recognising the long and variable lags in the transmission of monetary policy and the
inevitability of short-term shocks to price dynamics, policy-makers need to adopt a mediumterm orientation in their decisions. Attempts to fine-tune price developments on a short-term
basis are doomed to failure. Such attempts are likely to impart an excessive activism to
policy-making, such that interest rate decisions add volatility to the economy rather than
stabilise it.
In adopting an appropriate medium-term orientation, monetary policy-makers are accorded
an extra “degree of freedom” in setting the policy stance. There can be several paths of
interest rates consistent with the maintenance of price stability over the longer run. At least in
principle, the choice of a specific path can be used to support broader purposes, without
prejudicing the outlook for price stability over the medium term. This has been recognised by
the so-called flexible inflation targeting literature.11 This framework explicitly foresees the
10
See M.A. King (1997), “Changes in UK monetary policy: Rules and discretion in practice.” Journal of
Monetary Economics 39, pp. 81–97.
11
See L.E.O. Svennson (1998), “Inflation targeting in an open economy: Strict or flexible inflation targeting?”
Victoria Economic Commentaries 15(1).
5
use of monetary policy to smooth developments in economic activity over the business cycle,
while anchoring longer-term inflation expectations at levels consistent with price stability.
In principle, the flexibility accorded to policy-makers by the medium-term orientation of
monetary policy could be used in other directions, such as to contain financial imbalances or
limit asset price volatility. The feasibility and desirability of such alternative approaches rests
crucially on how monetary policy actions are perceived to be transmitted to the economy in
general, and the price level in particular – in other words, on the perception of the structure of
the economy and the monetary policy transmission mechanism.
2. The monetary policy transmission mechanism
The desirability of flexible inflation targeting was established using the canonical New
Keynesian model. In its simplest version, welfare is maximised by fully stabilising inflation
at its target level, a policy which simultaneously delivers a zero output gap. 12 This so-called
“divine coincidence”13 implies that policy-makers who care only about stabilising inflation
will deliver an efficient path of economic activity. And, symmetrically, policy-makers who
focus solely on keeping output at an (appropriately defined) potential level will deliver low
and stable inflation rates. A relatively simple Taylor-like monetary policy rule could achieve
these desirable outcomes. In this benign world, traditional monetary policy trade-offs appear
to have disappeared!
It is widely recognised that even modest deviations from the simplest version of the canonical
model introduce caveats to these very strong conclusions. But the underlying message of
these models remains: a policy aimed at stabilising inflation and the output gap at a horizon
around two years ahead is both desirable on welfare grounds and consistent with price
stability over the medium term.
Given our experience over the past few years, these strong conclusions are now rightly
treated with scepticism. With the benefit of hindsight, it is apparent they served to breed
complacency among both policy-makers and market participants, resulting in overconfidence in the ability of macro policies to stabilise the economy. Such complacency
contributed to both poor policy choices and destabilising private sector behaviour. Such
decisions may help to explain the emergence of financial crises in recent years.
12
At least when the output gap is defined in a “model-consistent” manner, i.e. as deviations of output from the
level it would achieve in the absence of nominal rigidities.
13
See: O. Blanchard and J. Galí (2007). “Real wage rigidities and the New Keynesian model,” Journal of
Money, Credit and Banking 39(s1), pp. 35-65.
6
Critiques of the benchmark New Keynesian model are legion. 14 I will not repeat them here.
Suffice to note that an important weakness of these models was the absence of a financial
sector. As a consequence, the financial intermediaries, financial frictions and asset price
bubbles and corrections that have played a key role in the events of the past three years were
neglected.
To illustrate, imagine a scenario where inflation falls because of a positive supply shock, such
as a reduction in manufacturing import prices from Emerging market economies. An
inflation-targeting central bank would decrease its policy rates to boost inflation back towards
its target level, simultaneously raising output closer to its new (temporarily higher) potential
level. In this scenario, financial market participants would experience both stronger economic
growth and lower interest rates, all in an environment of benign price developments. The
question arises of whether such circumstances can prompt financial market participants to
assume a euphoric state: to take on more risk; to expand balance sheets more rapidly; to
increase leverage; and to bid up asset prices. Such developments – by construction neglected
in the benchmark model lacking a financial sector – may have a macroeconomic impact and,
ultimately, undesirable consequences for both financial and price stability.
Why should low interest rates in these circumstances induce such euphoria in the financial
sector? And what implications can it have for credit conditions and the transmission of
monetary policy?
Recent experience suggests that the monetary policy stance may affect the confidence – and
thus the risk-taking behaviour – of financial institutions, in particular of banks.15 Since bank
loans are illiquid assets, liquidation is costly. When liquidity is tight, banks curtail their
accumulation of illiquid assets, notably by constraining credit supply. However, if the central
bank stands ready to lower interest rates when the financial system needs liquidity, financial
institutions may be induced to accumulate illiquid and riskier assets on their balance sheets,
given the implicit insurance provided by monetary policy. 16 Low policy interest rates may
also promote a “search for yield,” as investors seek to maintain their nominal rate of return at
14
Prominent examples include, inter alia: P. Krugman (2009), “How did economists get it so wrong?” New
York Times (2 September); and W.H. Buiter (2009), “The unfortunate uselessness of most ‘state-of-the-art’
academic monetary economics,” Financial Times (3 March).
15
See: C.E.V. Borio and H. Zhu (2008), “Capital regulation, risk-taking and monetary policy: A missing link
in the transmission mechanism?” BIS working paper no. 268.
16
See: Y. Altumbas, L. Gambacorta and D. Marques (2010), “Does monetary policy affect bank risk-taking?”
ECB working paper no. 1166; G. Jiménez, S.R.G. Ongena, J-L. Peydro and J. Saurina Salas (2010), “Credit
supply: Identifying balance sheet channels with loan applications and granted loans,” CEPR discussion
paper no. 7655; and A. Maddaloni and J-L. Peydró (2009), “Bank risk-taking, securitisation, supervision and
low interest rates: Evidence from lending standards,” ECB working paper, forthcoming.
7
historical levels, even as short risk-free rates decline. As interest rates fall, intermediaries take
on greater risk so as to meet this demand for returns.
Similar institutions will adopt a similar risk-taking strategy. Herding will emerge. Risk
management behaviour will become more correlated across financial intermediaries. Risk
will rise not only at the individual bank level, but also across the system as a whole.17
Moreover, since the financial sector is now exposed to greater liquidity risk as a result of
such behaviour, central banks are more likely to be called upon to intervene in the future.
And the magnitude of such interventions will need to increase if they are to be effective. A
“ratcheting-up” of risk within the banking sector can occur over time, creating increasing
vulnerability.18
Low interest rates also encourage a higher level of leverage. Because financial intermediaries
– whether banks, broker-dealers, shadow banks, or hedge funds – finance themselves with
short-term liabilities, central bank decisions on the level of short-term rate affect their
marginal price of leverage. As a result, the behaviour of financial intermediaries – which, by
their nature, have high levels of leverage – can be strongly influenced by even small changes
in short-term rates.19 Low levels of interest rates encourage higher levels of leverage and, as a
result, an accumulation of additional risk on bank balance sheets.
Through all these channels, a prolonged period of low short-term interest rates can support
the accumulation of risks and financial imbalances on the balance sheets of both financial
intermediaries and the private sector. These imbalances render the economy vulnerable to
financial crises: if confidence evaporates, lenders call in loans, balance sheets contract and a
painful readjustment is required, with adverse implications for financial stability and the real
economy. In particular, the build-up of leverage within the financial sector may unwind
abruptly, leading to a tightening of overall financing conditions.20
Such thinking sheds light on recent developments in the run-up to the financial crisis.
Viewing the impact of globalisation on goods prices as the manifestation of a positive supply
17
See: E. Fahri and J. Tirole (2009), “Leverage and the central banker's put,” American Economic Review
99(2), pp. 589-93.
18
See: D.W. Diamond and R.G. Rajan (2009), “Illiquidity and interest rate policy,” NBER working paper no.
15197.
19
See: T. Adrian and H.S. Shin (2009), “Financial intermediaries and monetary economics,” Handbook of
Monetary Economics, forthcoming; R.G. Rajan (2005), “Has financial development made the world riskier?”
Proceedings of the FRB Kansas City Jackson Hole symposium ‘The Greenspan years: Lessons for the
future’; and F. Allen and D. Gale (2007), Understanding financial crises, Oxford University Press.
20
See: G.B. Gorton and A. Metrick (2009), “Securitized banking and the run on repo,” Yale ICF working
paper no. 09-14.
8
shock, we can map this abstract scenario into our experience of the past decade. At the global
level, the incorporation of emerging markets into the world economy weighed down on price
developments (especially of manufactured goods). In response to the resulting benign
inflation developments, the prevailing inflation targeting orthodoxy led to the adoption of an
accommodative monetary policy stance. In retrospect, such accommodation was probably
inappropriate, at least to the extent that it supported the financial imbalances and asset price
bubbles that underlay the recent financial crisis. Failure to take into account financial factors
and channels of transmission led to a mis-calibration of the monetary policy response to the
impact of globalisation.
A simplistic account of the financial sector may have had other, adverse implications for the
conduct of monetary policy in recent years. Within the canonical New Keynesian model,
financial markets are “perfect.” On the basis of arbitrage considerations, long-term interest
rates and other yields and asset returns all derive from the expected path of short-term interest
rates determined by the central bank.
Such a construct suggests that monetary policy-makers can steer long-term rates closely by
simply making pre-commitments about the future path of short rates. No doubt there is an
element of truth in this characterisation. But the conditionality of such pre-commitments on
the state of the economy may prove hard to communicate. Financial intermediaries may
interpret pre-announced paths as a liquidity guarantee, promoting the excessive risk-taking
behaviour I have discussed. Such behaviour may then limit the future flexibility of monetary
policy-makers to react promptly to changed circumstances.
All in all, such pre-commitments are not useful and, may even be damaging. Effective central
bank communication should allow market participants to understand the decision-making
framework. But a pre-commitment to a future path of short-term rates is best avoided. It
limits the flexibility of policy-makers – operating within a transparent, rules-based
framework – to take decisions as conditions evolve. Looking back over recent years, it seems
likely that the adoption of such pre-commitments by some central banks led them to tighten
policy too late and too slowly, thus helping to maintain the overly loose monetary stance that
fuelled financial imbalances.
3. Assessing economic conditions
Even with the most sophisticated models of the economy, policy-makers are still heavily
reliant on an accurate assessment of the conjunctural situation. Yet the usual business cycle
9
indicators are subject to a variety of measurement errors. As has been demonstrated
extensively in the literature, the empirical proxies used to capture the output gap are subject
to constant revision.21 Clearly, policy-makers who rely exclusively on such assessments of
the cyclical position can be seriously misguided in their real-time conjunctural assessment.
To the extent that these indicators are central to the implementation of flexible inflation
targeting, over-reliance on them may have led to an over-activist and at times inappropriate
monetary policy stance. In particular, policy makers might also have formed an overly benign
assessment of the situation prevailing in 2002-2004 if they relied excessively on output gap
measures, as these measures were based on overly optimistic assumptions about potential
growth. For instance, in early 2003 the IMF estimated the US output gap for that same year to
be slightly above 2%. In early 2006 the estimate was revised down by 1 percentage point and
one year later by another 1 percentage point. With the benefit of hindsight it turned out that in
2003 the output gap was basically nil (See figure 3). 22
Such considerations plague other composite indicators derived from the literature, such as the
natural real interest rate.
Understanding current inflation developments can also prove challenging. Headline inflation
rates may be volatile in the short term, reflecting the impact of temporary relative price
shocks.23 Monetary policy-makers need to look through such noise. But conventional
measures of so-called “core inflation” can themselves be misleading. For example, measures
that exclude energy prices can offer a false impression of the underlying dynamics of
inflation if energy prices are trending upwards.
Clearly, policy-makers with a mistaken view of the conjunctural situation are liable to make
policy mistakes. In a stylised Taylor-rule setting, errors in assessing inflation or the output
gap will lead to flawed guidance on interest rate decisions. Such considerations may have led
to overly accommodative monetary policy prior to the financial crisis.
To give one example, I have already discussed how the forces of globalisation at the turn of
the century weighed on the prices of manufactured goods. This was an important factor
underpinning the benign outlook for inflation. Yet the rapid growth of emerging market
economies had other implications for global price developments. Notably, commodity prices
21
22
23
See: A. Orphanides and S. van Norden (2002), “The unreliability of output gap estimates in real time,”
Review of Economics and Statistics 84(4), pp. 569-583; and A. Orphanides and S. van Norden (2005), “The
reliability of inflation forecasts based on output gap estimates in real time,” Journal of Money, Credit and
Banking 37(3), pp. 583-601.
L. Bini Smaghi (2009), “Monetary Policy and Asset Prices”, Freiburg University, 14 October 2009.
Such as seasonal movements in food prices.
10
in general – and that of oil in particular – rose rapidly in the face of strong demand from
China as well as from Brazil, Russia and India. Policy-makers who focused on measures of
core inflation that excluded energy prices neglected the latter effect. They may have formed
an overly benign impression of the consequences of globalisation for price developments, and
adopted an overly loose monetary policy as a result.
Thus far, my discussion of the conjunctural assessment has focused on indicators motivated
by the benchmark New Keynesian framework. But we need to include the financial sector.
What indicators were missed by central banks prior to the crisis due to the neglect of financial
factors? What additional indicators should be embodied in the conjunctural assessment?
Monetary and credit aggregates can – and did – have a role to play. At least in part, the
ECB’s oft-maligned monetary analysis was vindicated by the crisis. If central banks had
taken the signals coming from strong monetary and credit growth in mid-decade more
seriously, monetary policy decisions might have been better calibrated to contain the
financial imbalances that ultimately led to the financial crisis.
But recent events have also revealed weaknesses in traditional monetary indicators.
Experience points to a need to deepen and broaden such analyses to take better account of the
fundamental changes observed in the financial sector over recent decades. New aggregates
and indicators need to be developed.24 Work at the ECB is proceeding in this direction.
In particular, the financial crisis revealed weakness in central banks’ understanding of
financial fragilities and their implications for the real economy and, ultimately, price
developments. To be concrete: while demonstrating concern at the rapid pace of credit
growth, in retrospect we did not understand well enough the accumulation of leverage in the
economy, in general – and in the financial system, in particular. And we took too benign a
view of the accelerating securitisation and credit risk transfer processes. Rather than
distributing risks to those best able to bear them, with the benefit of hindsight, incentive
problems in these processes led to a concentration of liquidity and credit risk.
In short, central banks had a poor understanding of the complex interconnections among
financial intermediaries and with the real economy, both domestically and internationally.
Systemic risk, contagion and herding behaviour were neglected. Addressing these
weaknesses in existing policy frameworks is at the heart of the ongoing development of socalled macro-prudential policy.
24
See T. Adrian and H.S. Shin (2009), “Prices and quantities in the monetary policy transmission mechanism,”
International Journal of Central Banking 5(4), pp. 131-142.
11
To sum up, the main weakness of monetary policy frameworks prior to the financial crisis
was their neglect of financial factors in the evolution of the economy and policy transmission.
We need to refine our policy frameworks to take account of such channels.
4.
Monetary policy and macro-prudential policy
Taking the role of the financial sector in monetary policy transmission more seriously reveals
the macroeconomic consequences of microeconomic behaviour: the impact on inflation,
growth and financial stability of the behaviour of financial institutions. As a result, a revised
framework has important implications for both the conduct of monetary policy and for
policies aimed at maintaining financial stability.
Viewing recent experience through this lens demonstrates that the achievement of financial
stability relies on more than traditional micro-prudential supervision of individual financial
institutions. Clearly, policy needs to take into account the impact of such externalities in
order to develop a framework supporting the stability of the financial system as a whole,
rather than the individual institutions which constitute it.
By the same token, the preceding analysis demonstrates that taking appropriate monetary
policy decisions relies on an understanding of the behaviour of financial intermediaries and,
in turn, on financial structure and innovation.
From a policy perspective, the impact of monetary policy on the behaviour of the financial
sector suggests that short-term interest rates are a potentially powerful tool to influence the
evolution of systemic risk, and thus to support financial stability. However, as I discussed at
the outset, monetary policy should not be overburdened with additional objectives. The
primacy of the price stability objective must be retained.
The well-known Tinbergen principle makes clear that a single instrument (short-term interest
rates) is insufficient to achieve two goals simultaneously.25 In such a context, additional
policy instruments are required. This is where macro-prudential policy tools come into play.
While the analytic basis and calibration of instruments such as pro-cyclical capital
requirements or leverage ratios require further elaboration, the hope is that such measures can
contain the accumulation of financial imbalances and vulnerabilities without recourse to
changes in short-term interest rates.
25
See: J. Tinbergen (1952), On the theory of economic policy, Amsterdam, North-Holland.
12
Monetary policy and macro-prudential policy: two policy instruments and two policy
objectives. At least in principle, the Tinbergen problem is resolved.
But practicalities stand in the way. As I said earlier, it is immediately apparent that monetary
policy and macro-prudential policy will interact in a variety of potentially complex ways,
throwing up new challenges for policy-making. Of course, what is required is the “right
combination” of prudential policies and monetary policy – a combination that simultaneously
achieves both price stability and financial stability. How is this combination best achieved?
We can draw some important lessons from our assessment of the interaction between
monetary and fiscal policy in the past. Attempts to develop an optimal “macroeconomic
policy mix” are superficially attractive, but have typically foundered in practice. As the
responsibilities of different policy-makers become blurred, their incentive to act appropriately
is diluted and ultimately the overall coherence of the policy stance is lost.
The institutional framework for monetary policy in the euro area reflects these concerns. The
single monetary policy is independent and has been assigned an unambiguous objective of
price stability. The ECB is accountable for the achievement of this objective. All this is
understood by other policy-makers. By acting in a transparent way consistent with its
mandate, the ECB creates an environment of price stability within which other authorities can
take decisions under their responsibility in order to achieve their own objectives. Clarity of
responsibility, independence of action and accountability for decisions in an environment of
open and frank exchange of information among policy authorities produces the best results.
These principles can be readily applied to the relationship between monetary policy and
macro-prudential policies. Of course, monetary policy decisions can have implications for
financial stability. And the degree of freedom accorded by the medium-term orientation of
monetary policy can be used to contain excessive risk-taking by banks. In this sense,
monetary policy can support financial stability objectives without prejudice to its primary
objective of price stability.
But this does not imply that monetary policy should be held jointly accountable for
maintaining financial stability, still less that it be assigned an additional financial stability
objective. Such measures would only serve to obscure the accountability of monetary policymakers for price stability and dilute the accountability of those authorities responsible for
financial stability. All this leads me to re-iterate my earlier conclusion: the primary objective
of monetary policy must remain price stability.
13
By contrast, the institutional structure for prudential supervision and regulation at the
European level remains at an embryonic stage, especially on the macro-prudential side. The
prospective creation of a European Systemic Risk Board will partly fill this lacuna.
For the reasons I have discussed, financial supervision and regulation clearly cannot be
conducted without reference to the monetary policy stance. But, both in finalising the
institutional framework and in the conduct of these policies in the future, the principles I have
articulated must be respected. We need to ensure that macro-prudential policy is formulated
both independently and transparently, guaranteeing a rich flow of information among the
relevant authorities, including monetary policy-makers, while at the same time avoiding any
blurring of responsibilities, objectives and accountability.
Concluding remarks
In conclusion, what are the main lessons from recent experience that we should seek to
incorporate in any revision to the framework for monetary policy-making?
First and foremost, monetary policy should have remained more closely focused on the
maintenance of price stability over the medium term. This implies that this objective must be
defined clearly and pursued symmetrically. Monetary policy should not be burdened with
additional objectives, which it is ill-equipped to pursue.
The institutional independence of the central bank is essential to building the credibility
required to pursue price stability in a consistent and coherent manner. Dual mandates for
monetary policy place this independence at risk. Independence accords central banks the
necessary flexibility to deal with a rapidly changing world without putting their credibility at
risk.
Second, monetary policy should have been less geared to fine-tuning the economy, in
particular to trying to reduce the output gaps which ex post turn out to be very different from
their real-time measurement. The long and variable lags in monetary policy transmission
mean that a medium-term orientation for monetary policy must be maintained. Temporary
deviations from a precise inflation objective are inevitable – monetary policy-makers must
focus on containing persistent trends in inflation. Distinguishing between temporary and
more persistent shocks to inflation is therefore crucial. Experience has shown that paying
attention to specific measures of “core inflation” may mislead monetary policy. An overly
mechanical view of such indicators can lead to an underestimation of the strength of
inflationary pressures at the global level.
14
Third, policy decisions should have been based on better models of monetary policy
transmission. Monetary and financial factors have been too easily dismissed, especially by
inflation targeting regimes. Placing greater weight on monetary and credit indicators should
allow interest rate decisions to be better calibrated to achieve the appropriate medium-term
objectives of monetary policy.
Can central banks learn from experience? They can if they are able to recognise what went
wrong, rather than sweeping the hard questions under the carpet, and if they are able to adapt
their analytical and decision-making framework. Suggesting that monetary policy had
nothing to do with the crisis will not help, and might encourage us to make the same mistakes
in the future. The reaction to the crisis has however shown that central banks learn fast and
can take decisive action to protect the economy. Their response is in stark contrast to how
they reacted to the Great Depression. On that occasion, monetary policy-makers were largely
responsible for causing the crisis, for deepening it and for preventing a quick recovery.
We can be more optimistic this time around.
15
Figure 1: Overall inflation
US
OECD countries
World
6.0
5.5
5.0
4.5
4.0
3.5
3.0
2.5
2.0
2003
2004
2005
2006
2007
2008
Source: ECB, BLS, IMF and OECD
Note: Average annual inflation rates
Figure 2: Overall inflation excluding food and energy
US
OECD countries
2.5
2.0
1.5
1.0
2003
2004
2005
2006
2007
2008
Source: ECB, BLS and OECD
Note: Average annual inflation rates
16
Figure 3
Spring vintages (2002-2009) of US output gap estimates by the IMF (WEO)
Apr-02
Apr-03
Apr-04
Apr-05
Apr-06
Apr-07
Apr-08
Apr-09
6.0
6.0
4.0
4.0
2.0
2.0
0.0
0.0
-2.0
-2.0
-4.0
-4.0
-6.0
-6.0
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Source: IMF (WEO) data
Note: Output gaps are defined as the percentage deviation of actual output from potential output.
17